Zabbix动态监控磁盘I/O

需要监控每台服务器的磁盘I/O，由于每台服务器的磁盘名称可能不一样，所以需要用到Zabbix的LLD功能来发现磁盘并监控I/O。

一、客户端（被监控端）

1.发现磁盘的脚本（disk_discovery.sh）

#!/bin/bash

#written by lenwood

#mail:

diskarray=(`cat /proc/diskstats |grep -E "\b[sv]d[abcdefg]\b|\bxvd[abcdefg]\b"|grep -i "\b$1\b"|awk '{print $3}'|sort|uniq   2>/dev/null`)

length=${#diskarray[@]}

printf "{\n"

printf  '\t'"\"data\":["

for ((i=0;i<$length;i++))

do

printf '\n\t\t{'

printf "\"{#DISK_NAME}\":\"${diskarray[$i]}\"}"

if [ $i -lt $[$length-1] ];then

printf ','

fi

done

printf  "\n\t]\n"

 

printf "}\n"

我把这个脚本放到zabbix-agent的/etc/zabbix/os目录下

执行结果如下：

root@ubuntu:~# ./disk_discovery.sh
{
	"data":[
		{"{#DISK_NAME}":"sda"}
	]
}

说明：执行结果看到此台测试机上面只有一块磁盘sda

2.获取磁盘I/O信息

a.使用iostat收集磁盘I/O信息（这条指令是每隔一秒将磁盘信息内容写入到/tmp/iostat_output,并且后台运行）

 1.apt-get install sysstat（有些机器上面没有安装sysstat，如果有了可以忽略此步骤）

  安装完之后可以使用 iostat 展示信息如下：

root@ubuntu:~# iostat
Linux 4.4.0-21-generic (ubuntu) 	10/28/2019 	_x86_64_	(1 CPU)

avg-cpu:  %user   %nice %system %iowait  %steal   %idle
           0.74    0.00    0.55    0.74    0.00   97.97

Device:            tps    kB_read/s    kB_wrtn/s    kB_read    kB_wrtn
sda               2.68         8.19        52.80    1383496    8915276

root@ubuntu:~#

Zabbix官方动态监控磁盘I/O

安装
在zabbix-agent.conf中增加参数

#### enable Include in config
# /etc/zabbix/zabbix_agentd.conf
echo "Include=/etc/zabbix-agent.d/*.conf" >> /etc/zabbix/zabbix_agentd.conf

#### copy config 
mkdir -p /etc/zabbix-agent.d/ && cd /etc/zabbix-agent.d/
curl -O 'https://raw.githubusercontent.com/dusharu/ZBX_Disk_Stat/master/config_zabbix-agent/Disk_Stat.conf'

#### restart zabbix-agent
# RHEL/CentOS
systemctl restart zabbix-agentd
# Debian/Ubuntu
systemctl restart zabbix-agent
# Gentoo
/etc/init.d/zabbix-agentd restart

Zabbix官方模板

wgethttps://github.com/dusharu/ZBX_Disk_Stat/blob/master/Template_ZBX/ZBX_Disk_Stat.xml

调试命令

测试发现

zabbix_get -s <IP|Server_Name> -k custom.blkdev.discovery |jq .

获得统计量

zabbix_get -s <IP|Server_Name> -k custom.blkdev.all_stat[vg00-lv_root] | jq .

监控磁盘I/O官方脚本
https://github.com/dusharu/ZBX_Disk_Stat/blob/master/Develop/scripts/blkdev_discovery_awk.sh
https://github.com/dusharu/ZBX_Disk_Stat/blob/master/Develop/scripts/blkdev_get_param_awk.sh
https://github.com/dusharu/ZBX_Disk_Stat/blob/master/Develop/scripts/blkdev_show_stat.sh
官方监控模板
https://github.com/dusharu/ZBX_Disk_Stat/blob/master/Template_ZBX/ZBX_Disk_Stat.xml

官方监控配置文件Disk_Stat.conf

##### ZBX_Disk_Stat #####
# Downloaded from:
#   https://github.com/dusharu/ZBX_Disk_Stat
#
# Docs:
#   https://www.kernel.org/doc/Documentation/admin-guide/iostats.rst 
#   https://www.kernel.org/doc/Documentation/ABI/testing/procfs-diskstats

#### Discovery ####
UserParameter=custom.blkdev.discovery,awk ' BEGIN { printf("{\n\t\"data\":["); } { if (FNR != 1) {printf ("\t},")}; printf("\n\t\t{\t\"{#DEVBLK}\":\""); if (match($3, /^dm-[0-9]*/)) { system("ls -l /dev/mapper/ |grep -e \"../" $3  "$\" | awk \"{print \\$9}\"|tr -d \"\\n\""); } else { printf($3); }; printf("\""); } END { {printf ("\t}\n")} printf("\t]\n}\n"); }' /proc/diskstats

#### Get Stat ####
UserParameter=custom.blkdev.all_stat[*],DISK="$1"; if [[ ! -b "/dev/$DISK" ]]; then DISK="$( ls -l /dev/mapper/ |awk "/ $DISK / {gsub(\"../\",\"\",\$$11);print \$$11}" )"; fi; if [[ "$DISK" != "" ]]; then grep -m1 " $DISK " /proc/diskstats | column --json --table-name ZBX_Disk_Stat --table-columns major,minor,name,read_complete,read_merge,read_sector,read_time,write_complete,write_merge,write_sector,write_time,io_queue,io_time,io_time_weight,discard_complete,discard_merge,discard_sector,discard_time --table-hide  major,minor,read_merge,write_merge,io_time,io_time_weight,discard_complete,discard_merge,discard_sector,discard_time; else exit 100; fi

地址：https://github.com/dusharu/ZBX_Disk_Stat/blob/master/config_zabbix-agent/Disk_Stat.conf

参考链接

https://www.iteye.com/blog/20120923lina-2258119
https://github.com/dusharu/ZBX_Disk_Stat